This is the mail archive of the cygwin mailing list for the Cygwin project.

Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: request for /dev/stdin and friends

On Nov 16 16:25, Eric Blake wrote:
> Corinna Vinschen <corinna-cygwin <at>> writes:
> > > > So what we should do here IMO is to augment the base-files package
> > > > to create a /dev directory and create a couple of standard symlinks
> > > > in it, which are not covered automatically by the Cygwin DLL.
> > > 
> > > There already was some discussion, ending in
> > > <>, which
> > > never got a reply...
> > 
> > Isn't your create_devices script adding all the devices handled by
> > Cygwin?  The idea here is slightly different in that the script should
> > only create symlinks in /dev.  Device nodes for the exising Cygwin
> > devices are not actually necessary so we should be careful with them,
> > maybe.
> One of my concerns is that readdir() on /dev works nicely, integrating both 
> symlinks and actual cygwin devices.  Right now, the benefit of Igor's script 
> creating placeholders even for actual devices is that readdir() (and thus 
> ls /dev, or even tab completion when you type /dev/<TAB>) can see all devices 
> and not just symlinks like /dev/stdin.

Yeah, right.  I'm thinking along the same lines.  Consider the situation
that /dev/ is handled automatically at one point in aa udev sort of way.
In that case static device nodes would be in the way.  Or, hmm, maybe not.
But that's why I suggested to be careful for now.


Corinna Vinschen                  Please, send mails regarding Cygwin to
Cygwin Project Co-Leader          cygwin AT cygwin DOT com
Red Hat

Unsubscribe info:
Problem reports:

Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]